Tsunami of April 1, 1946 — Macdonald et al. 
25 
eating an average interval between early wave 
crests of approximately 12.5 minutes. The 
interval between the first wave crest and the 
succeeding trough was 7.5 minutes, however, 
indicating a wave period of 15 minutes at 
the beginning of the disturbance. This cor¬ 
responds with the mean wave period of 15.6 
minutes found by Green (1946: 499) at 
Honolulu and eight other stations on the 
coasts of North and South America. At the 
mouth of Nuuanu Stream in Honolulu, C. K. 
Wentworth observed an interval of approxi¬ 
mately 15 minutes between successive bores 
ascending the stream, and a wave period of 
about 15 minutes was observed by J. B. Cox 
and D. C. Cox at Waikiki at about 7:45 a.m. 
Observations elsewhere were poor, but in 
general indicated an interval not far from 
15 minutes between the early waves of the 
series. The interval between later waves at 
Honolulu (Fig. 2) and elsewhere was 
shorter and less regular, probably because of 
the arrival of chains of waves traveling by 
somewhat different routes, refracted around 
different sides of islands, and reflected at 
various points, as well as traveling by the 
most direct route. Probably contributing to 
the irregularity of later waves were wind 
waves and also the free-period oscillations, 
in harbors and channels, known as "seiches.” 
If the period of the waves is assumed to be 
15 minutes, and the average speed to be 
about 489 miles an hour, the average wave 
length from crest to crest was about 122 
miles. 
Direct observations on the height of the 
waves in the open sea are lacking, but theo¬ 
retical considerations indicate that the height 
probably did not exceed 2 feet from crest 
to trough. 4 If so, the small height combined 
with the very great wave length should have 
made the waves imperceptible to ships at sea. 
That such was indeed the case is indicated 
4 Based on the assumption of a 10-foot wave in 
10 feet of water, and the variation of the wave 
height inversely as the fourth root of the depth. 
by the fact that the master of a ship lying 
offshore near Hilo could feel no unusual 
waves, although he could see the great waves 
breaking onshore. Crews of fishing boats in 
the Hawaiian area also reported no unusual 
conditions at the time of the tsunami, al¬ 
though heavy storm waves were running. 
The few reports of violent waves of great 
height from ships at sea were probably occa¬ 
sioned by storm waves, together with the 
knowledge that a tsunami was taking place. 
The nature of the waves sweeping up on 
to Hawaiian shores varied greatly from place 
to place. At some places the water rose 
gently, flooding over the coastal lands with¬ 
out the development of any steep wave front. 
At such places most of the damage resulted 
from the violent run-back of the water to the 
sea. At some localities, although the gen¬ 
eral water surface rose gently, ordinary storm 
waves moved in over the top of the broad 
swells of the tsunami, and there at least part 
of the damage was caused by the storm 
waves. At most places, however, the waves 
of the tsunami swept toward shore with 
steep fronts and great turbulence, causing a 
loud roaring and hissing noise. Locally, the 
wave closely resembled a tidal bore, the steep 
front rolling in over comparatively quiet 
water in front of it. Behind the steep front, 
the wave crest was broad and nearly flat, 
with smaller storm waves superimposed upon 
it. Such bores were best developed in bays 
and estuaries, but waves of closely similar 
form were observed crossing shallowly sub¬ 
merged reefs upon otherwise open coasts. 
At many places the violence of the waves 
moving shoreward was sufficiently great to 
tear loose heads of coral and algae, up to 4 
feet across, and toss them onto the beach 
as much as 15 feet above sea level. Locally, 
blocks of reef rock weighing several tons 
were quarried at the outer edge of the reef 
and thrown onto the reef surface. 
Between crests, the water withdrew from 
shore, exposing reefs, coastal mud flats, and 
